deep lang.mixin
Locked
 
|
I opened an enhancement request (https://bugs.dojotoolkit.org/ticket/18958) today but probably should have started here first...
Basically I would like a deep mixin similar to jQuey's extend(true, obj1, obj2) overload... var object1 = { apple: 0, banana: { weight: 52, price: 100 }, cherry: 97 }; var object2 = { banana: { price: 200 }, durian: 100 }; // Merge object2 into object1, recursively $.extend( true, object1, object2 ); //object1 = {"apple":0,"banana":{"weight":52,"price":200},"cherry":97,"durian":100} I have made a sample Plunker with jQuery.extend brute force converted to lang.jmixin using dojo's isObject, IsArray and isFunction in app/lang.js which is a copy of current dojo/_base/lang. It works for the above input, but as dojo doesn't have a function similar to isPlainObject that I can see, it may not work exactly as theirs does. Plunker link: http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview -- Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest |
|
|
Hi Michael,
It's in an odd location, but I believe what you are looking for is the same as dojo/request/util.deepCopy See https://jsfiddle.net/dylan/oja45dv8/ Regards, -Dylan on 1/17/17, 21:06 (GMT-07:00) Michael Schall said the following: > I opened an enhancement request > (https://bugs.dojotoolkit.org/ticket/18958) today but probably should > have started here first... > > Basically I would like a deep mixin similar to jQuey's extend(true, > obj1, obj2) overload... > > var object1 = { > apple: 0, > banana: { weight: 52, price: 100 }, > cherry: 97 > }; > var object2 = { > banana: { price: 200 }, > durian: 100 > }; > > // Merge object2 into object1, recursively > $.extend( true, object1, object2 ); > > //object1 = > {"apple":0,"banana":{"weight":52,"price":200},"cherry":97,"durian":100} > > I have made a sample Plunker with jQuery.extend brute force converted to > lang.jmixin using dojo's isObject, IsArray and isFunction in app/lang.js > which is a copy of current dojo/_base/lang. It works for the above > input, but as dojo doesn't have a function similar to isPlainObject that > I can see, it may not work exactly as theirs does. > > Plunker link: http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview > Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest
Co-Founder, Dojo Toolkit
CEO, SitePen, Inc. http://www.sitepen.com/ |
|
|
Dylan - Thanks for the quick response! I tested this today a bit and I think this may work for our cases. Looks like dojo 2's core/lang exports a deepMixin function ( https://github.com/dojo/core/blob/master/src/lang.ts#L165 ) that will support this in the future and has special handling for arrays similar to the jQuery implementation? Could core/lang's deepMixin be back ported to 1.x's lang module? Mike On Wed, Jan 18, 2017 at 6:32 AM, Dylan Schiemann <[hidden email]> wrote: Hi Michael, -- Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest |
|
|
Hi Mike,
We could. I'd be happy to review a PR if someone wants to backport it. Another option of course could be to use @dojo/core within your Dojo 1 app. It would be larger than just adding a function to dojo/_base/lang.js, but @dojo/core is pretty small. I'd be curious to hear how that experience goes if you're interested in that experiment. You could basically transpile @dojo/core and then use @dojo/core/lang as an AMD module. Regards, -Dylan on 1/18/17, 17:31 (GMT-07:00) Michael Schall said the following: > Dylan - Thanks for the quick response! > > I tested this today a bit and I think this may work for our cases. > > Looks like dojo 2's core/lang exports a deepMixin function ( > https://github.com/dojo/core/blob/master/src/lang.ts#L165 ) that will > support this in the future and has special handling for arrays similar > to the jQuery implementation? > > Could core/lang's deepMixin be back ported to 1.x's lang module? > > Mike > > On Wed, Jan 18, 2017 at 6:32 AM, Dylan Schiemann <[hidden email] > <mailto:[hidden email]>> wrote: > > Hi Michael, > > It's in an odd location, but I believe what you are looking for is the > same as dojo/request/util.deepCopy > > See https://jsfiddle.net/dylan/oja45dv8/ > <https://jsfiddle.net/dylan/oja45dv8/> > > Regards, > -Dylan > > on 1/17/17, 21:06 (GMT-07:00) Michael Schall said the following: > > I opened an enhancement request > > (https://bugs.dojotoolkit.org/ticket/18958 > <https://bugs.dojotoolkit.org/ticket/18958>) today but probably should > > have started here first... > > > > Basically I would like a deep mixin similar to jQuey's extend(true, > > obj1, obj2) overload... > > > > var object1 = { > > apple: 0, > > banana: { weight: 52, price: 100 }, > > cherry: 97 > > }; > > var object2 = { > > banana: { price: 200 }, > > durian: 100 > > }; > > > > // Merge object2 into object1, recursively > > $.extend( true, object1, object2 ); > > > > //object1 = > > > {"apple":0,"banana":{"weight":52,"price":200},"cherry":97,"durian":100} > > > > I have made a sample Plunker with jQuery.extend brute force > converted to > > lang.jmixin using dojo's isObject, IsArray and isFunction in > app/lang.js > > which is a copy of current dojo/_base/lang. It works for the above > > input, but as dojo doesn't have a function similar to > isPlainObject that > > I can see, it may not work exactly as theirs does. > > > > Plunker link: http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview > <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview> > > > -- > Dojo Toolkit: http://dojotoolkit.org/ > Tutorials: http://dojotoolkit.org/documentation/ > <http://dojotoolkit.org/documentation/> > > [hidden email] > <mailto:[hidden email]> > To unsubscribe, visit: > http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest > <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> > > Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest
Co-Founder, Dojo Toolkit
CEO, SitePen, Inc. http://www.sitepen.com/ |
|
|
I agree pulling in parts of Dojo 2 would be interesting, but probably don't want to pull that much in for a single function... When I get time to look into pulling this back, would the correct end goal be: Backward compatibility: only add a deepMixin (and renamed private called functions as they collide) and leave current mixin (and called functions) unchanged? Smaller file: pull back both deepMixin and mixin? The smaller file approach would probably be best, but I'm not sure the mixin functions between 1.x and 2.x are functionally equivalent. The tests I can find for dojo 1.x lang.mixin (https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js#L67) seem pretty sparse as compared to the ones for 2.x (https://github.com/dojo/core/blob/master/tests/unit/lang.ts#L142). On Thu, Jan 19, 2017 at 6:44 AM Dylan Schiemann <[hidden email]> wrote: Hi Mike, -- Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest |
|
|
I see a few approaches:
1. Add to dojo/_base/lang. Probably the simplest, but limits what you can do. 2. Add a new dojo/mixin module. Ok, but would probably add confusion. 3. Add a new dojox/lang/mixin module. Probably the best place to add it, even though it might not get a lot of attention there. At least then you could add better versions of mixin and deepMixin and not risk confusion with what is in dojo/_base/lang So I'd probably lean towards #3 as the right approach given the constraints. Regards, -Dylan on 1/19/17, 11:18 (GMT-07:00) Michael Schall said the following: > I agree pulling in parts of Dojo 2 would be interesting, but probably > don't want to pull that much in for a single function... > > When I get time to look into pulling this back, would the correct end > goal be: > > Backward compatibility: only add a deepMixin (and renamed private called > functions as they collide) and leave current mixin (and called > functions) unchanged? > Smaller file: pull back both deepMixin and mixin? > > The smaller file approach would probably be best, but I'm not sure the > mixin functions between 1.x and 2.x are functionally equivalent. The > tests I can find for dojo 1.x lang.mixin > (https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js#L67 > <https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js>) > seem pretty sparse as compared to the ones for 2.x > (https://github.com/dojo/core/blob/master/tests/unit/lang.ts#L142). > > > On Thu, Jan 19, 2017 at 6:44 AM Dylan Schiemann <[hidden email] > <mailto:[hidden email]>> wrote: > > Hi Mike, > > We could. I'd be happy to review a PR if someone wants to backport it. > > Another option of course could be to use @dojo/core within your Dojo 1 > app. It would be larger than just adding a function to > dojo/_base/lang.js, but @dojo/core is pretty small. I'd be curious to > hear how that experience goes if you're interested in that experiment. > You could basically transpile @dojo/core and then use @dojo/core/lang as > an AMD module. > > Regards, > -Dylan > > on 1/18/17, 17:31 (GMT-07:00) Michael Schall said the following: > > Dylan - Thanks for the quick response! > > > > I tested this today a bit and I think this may work for our cases. > > > > Looks like dojo 2's core/lang exports a deepMixin function ( > > https://github.com/dojo/core/blob/master/src/lang.ts#L165 ) that will > > support this in the future and has special handling for arrays similar > > to the jQuery implementation? > > > > Could core/lang's deepMixin be back ported to 1.x's lang module? > > > > Mike > > > > On Wed, Jan 18, 2017 at 6:32 AM, Dylan Schiemann > <[hidden email] <mailto:[hidden email]> > > <mailto:[hidden email] <mailto:[hidden email]>>> wrote: > > > > Hi Michael, > > > > It's in an odd location, but I believe what you are looking > for is the > > same as dojo/request/util.deepCopy > > > > See https://jsfiddle.net/dylan/oja45dv8/ > > <https://jsfiddle.net/dylan/oja45dv8/> > > > > Regards, > > -Dylan > > > > on 1/17/17, 21:06 (GMT-07:00) Michael Schall said the following: > > > I opened an enhancement request > > > (https://bugs.dojotoolkit.org/ticket/18958 > > <https://bugs.dojotoolkit.org/ticket/18958>) today but > probably should > > > have started here first... > > > > > > Basically I would like a deep mixin similar to jQuey's > extend(true, > > > obj1, obj2) overload... > > > > > > var object1 = { > > > apple: 0, > > > banana: { weight: 52, price: 100 }, > > > cherry: 97 > > > }; > > > var object2 = { > > > banana: { price: 200 }, > > > durian: 100 > > > }; > > > > > > // Merge object2 into object1, recursively > > > $.extend( true, object1, object2 ); > > > > > > //object1 = > > > > > > {"apple":0,"banana":{"weight":52,"price":200},"cherry":97,"durian":100} > > > > > > I have made a sample Plunker with jQuery.extend brute force > > converted to > > > lang.jmixin using dojo's isObject, IsArray and isFunction in > > app/lang.js > > > which is a copy of current dojo/_base/lang. It works for > the above > > > input, but as dojo doesn't have a function similar to > > isPlainObject that > > > I can see, it may not work exactly as theirs does. > > > > > > Plunker link: > http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview > > <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview> > > > > > -- > > Dojo Toolkit: http://dojotoolkit.org/ > > Tutorials: http://dojotoolkit.org/documentation/ > > <http://dojotoolkit.org/documentation/> > > > > [hidden email] > <mailto:[hidden email]> > > <mailto:[hidden email] > <mailto:[hidden email]>> > > To unsubscribe, visit: > > http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest > > <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> > > > > > -- > Dojo Toolkit: http://dojotoolkit.org/ > Tutorials: http://dojotoolkit.org/documentation/ > > [hidden email] > <mailto:[hidden email]> > To unsubscribe, visit: > http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest > Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest
Co-Founder, Dojo Toolkit
CEO, SitePen, Inc. http://www.sitepen.com/ |
|
|
Dylan - Thanks for your help and direction... I think I'm going to go back and look at your initial suggestion... If you think the "best" approach is to create a dojox package, why not just look into pulling in dojo 2 lang as a module. From the looks of it, lang is almost self contained. Do you know if there any examples/documentation of how to transpile/build dojo 2? How do you get it to support AMD? I've looked, but have only come across the build examples for dojo 1.x. On Thu, Jan 19, 2017 at 3:37 PM, Dylan Schiemann <[hidden email]> wrote: I see a few approaches: -- Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest |
|
|
Hi Mike,
The process for dojo/core would be to npm install core and its dependencies (has and shim): npm install @dojo/shim npm install @dojo/has npm install @dojo/core You may get a few spurious warnings. Then the packages should exist as UMD modules in node_modules/@dojo/ , so add the location of the @dojo package in your dojo config, and you should be good to go by requiring @dojo/core/lang as an AMD module. I have not tried building these modules with the Dojo 1 build system, but this version is transpiled to ES5, so things should just work hopefully. Regards, -Dylan on 1/23/17, 20:45 (GMT-07:00) Michael Schall said the following: > Dylan - > > Thanks for your help and direction... > > I think I'm going to go back and look at your initial suggestion... If > you think the "best" approach is to create a dojox package, why not just > look into pulling in dojo 2 lang as a module. From the looks of it, > lang is almost self contained. > > Do you know if there any examples/documentation of how > to transpile/build dojo 2? How do you get it to support AMD? I've > looked, but have only come across the build examples for dojo 1.x. > > On Thu, Jan 19, 2017 at 3:37 PM, Dylan Schiemann <[hidden email] > <mailto:[hidden email]>> wrote: > > I see a few approaches: > > 1. Add to dojo/_base/lang. Probably the simplest, but limits what you > can do. > 2. Add a new dojo/mixin module. Ok, but would probably add confusion. > 3. Add a new dojox/lang/mixin module. Probably the best place to add it, > even though it might not get a lot of attention there. At least then you > could add better versions of mixin and deepMixin and not risk confusion > with what is in dojo/_base/lang > > So I'd probably lean towards #3 as the right approach given the > constraints. > > Regards, > -Dylan > > on 1/19/17, 11:18 (GMT-07:00) Michael Schall said the following: > > I agree pulling in parts of Dojo 2 would be interesting, but probably > > don't want to pull that much in for a single function... > > > > When I get time to look into pulling this back, would the correct end > > goal be: > > > > Backward compatibility: only add a deepMixin (and renamed private > called > > functions as they collide) and leave current mixin (and called > > functions) unchanged? > > Smaller file: pull back both deepMixin and mixin? > > > > The smaller file approach would probably be best, but I'm not sure the > > mixin functions between 1.x and 2.x are functionally equivalent. The > > tests I can find for dojo 1.x lang.mixin > > (https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js#L67 <https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js#L67> > > <https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js > <https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js>>) > > seem pretty sparse as compared to the ones for 2.x > > (https://github.com/dojo/core/blob/master/tests/unit/lang.ts#L142 > <https://github.com/dojo/core/blob/master/tests/unit/lang.ts#L142>). > > > > > > On Thu, Jan 19, 2017 at 6:44 AM Dylan Schiemann > <[hidden email] <mailto:[hidden email]> > > <mailto:[hidden email] <mailto:[hidden email]>>> wrote: > > > > Hi Mike, > > > > We could. I'd be happy to review a PR if someone wants to > backport it. > > > > Another option of course could be to use @dojo/core within your > Dojo 1 > > app. It would be larger than just adding a function to > > dojo/_base/lang.js, but @dojo/core is pretty small. I'd be > curious to > > hear how that experience goes if you're interested in that > experiment. > > You could basically transpile @dojo/core and then use > @dojo/core/lang as > > an AMD module. > > > > Regards, > > -Dylan > > > > on 1/18/17, 17:31 (GMT-07:00) Michael Schall said the following: > > > Dylan - Thanks for the quick response! > > > > > > I tested this today a bit and I think this may work for our > cases. > > > > > > Looks like dojo 2's core/lang exports a deepMixin function ( > > > https://github.com/dojo/core/blob/master/src/lang.ts#L165 > <https://github.com/dojo/core/blob/master/src/lang.ts#L165> ) that will > > > support this in the future and has special handling for > arrays similar > > > to the jQuery implementation? > > > > > > Could core/lang's deepMixin be back ported to 1.x's lang module? > > > > > > Mike > > > > > > On Wed, Jan 18, 2017 at 6:32 AM, Dylan Schiemann > > <[hidden email] <mailto:[hidden email]> > <mailto:[hidden email] <mailto:[hidden email]>> > > > <mailto:[hidden email] <mailto:[hidden email]> > <mailto:[hidden email] <mailto:[hidden email]>>>> wrote: > > > > > > Hi Michael, > > > > > > It's in an odd location, but I believe what you are looking > > for is the > > > same as dojo/request/util.deepCopy > > > > > > See https://jsfiddle.net/dylan/oja45dv8/ > <https://jsfiddle.net/dylan/oja45dv8/> > > > <https://jsfiddle.net/dylan/oja45dv8/ > <https://jsfiddle.net/dylan/oja45dv8/>> > > > > > > Regards, > > > -Dylan > > > > > > on 1/17/17, 21:06 (GMT-07:00) Michael Schall said the > following: > > > > I opened an enhancement request > > > > (https://bugs.dojotoolkit.org/ticket/18958 > <https://bugs.dojotoolkit.org/ticket/18958> > > > <https://bugs.dojotoolkit.org/ticket/18958 > <https://bugs.dojotoolkit.org/ticket/18958>>) today but > > probably should > > > > have started here first... > > > > > > > > Basically I would like a deep mixin similar to jQuey's > > extend(true, > > > > obj1, obj2) overload... > > > > > > > > var object1 = { > > > > apple: 0, > > > > banana: { weight: 52, price: 100 }, > > > > cherry: 97 > > > > }; > > > > var object2 = { > > > > banana: { price: 200 }, > > > > durian: 100 > > > > }; > > > > > > > > // Merge object2 into object1, recursively > > > > $.extend( true, object1, object2 ); > > > > > > > > //object1 = > > > > > > > > > > {"apple":0,"banana":{"weight":52,"price":200},"cherry":97,"durian":100} > > > > > > > > I have made a sample Plunker with jQuery.extend brute > force > > > converted to > > > > lang.jmixin using dojo's isObject, IsArray and > isFunction in > > > app/lang.js > > > > which is a copy of current dojo/_base/lang. It works for > > the above > > > > input, but as dojo doesn't have a function similar to > > > isPlainObject that > > > > I can see, it may not work exactly as theirs does. > > > > > > > > Plunker link: > > http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview > <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview> > > > <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview > <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview>> > > > > > > > -- > > > Dojo Toolkit: http://dojotoolkit.org/ > > > Tutorials: http://dojotoolkit.org/documentation/ > <http://dojotoolkit.org/documentation/> > > > <http://dojotoolkit.org/documentation/ > <http://dojotoolkit.org/documentation/>> > > > > > > [hidden email] > <mailto:[hidden email]> > > <mailto:[hidden email] > <mailto:[hidden email]>> > > > <mailto:[hidden email] > <mailto:[hidden email]> > > <mailto:[hidden email] > <mailto:[hidden email]>>> > > > To unsubscribe, visit: > > > > http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest > <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> > > > > <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest > <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest>> > > > > > > > > -- > > Dojo Toolkit: http://dojotoolkit.org/ > > Tutorials: http://dojotoolkit.org/documentation/ > <http://dojotoolkit.org/documentation/> > > > > [hidden email] > <mailto:[hidden email]> > > <mailto:[hidden email] > <mailto:[hidden email]>> > > To unsubscribe, visit: > > http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest > <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> > > > -- > Dojo Toolkit: http://dojotoolkit.org/ > Tutorials: http://dojotoolkit.org/documentation/ > <http://dojotoolkit.org/documentation/> > > [hidden email] > <mailto:[hidden email]> > To unsubscribe, visit: > http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest > <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> > > Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest
Co-Founder, Dojo Toolkit
CEO, SitePen, Inc. http://www.sitepen.com/ |
|
|
P.S. https://github.com/dojo/widgets#usage has this documented, and will
soon be added to the READMEs for the other packages, as well as the tutorials which are starting to be authored. Regards, -Dylan on 1/24/17, 05:27 (GMT-07:00) Dylan Schiemann said the following: > Hi Mike, > > The process for dojo/core would be to npm install core and its > dependencies (has and shim): > > npm install @dojo/shim > npm install @dojo/has > npm install @dojo/core > > You may get a few spurious warnings. > > Then the packages should exist as UMD modules in node_modules/@dojo/ , > so add the location of the @dojo package in your dojo config, and you > should be good to go by requiring @dojo/core/lang as an AMD module. > > I have not tried building these modules with the Dojo 1 build system, > but this version is transpiled to ES5, so things should just work hopefully. > > Regards, > -Dylan > > on 1/23/17, 20:45 (GMT-07:00) Michael Schall said the following: >> Dylan - >> >> Thanks for your help and direction... >> >> I think I'm going to go back and look at your initial suggestion... If >> you think the "best" approach is to create a dojox package, why not just >> look into pulling in dojo 2 lang as a module. From the looks of it, >> lang is almost self contained. >> >> Do you know if there any examples/documentation of how >> to transpile/build dojo 2? How do you get it to support AMD? I've >> looked, but have only come across the build examples for dojo 1.x. >> >> On Thu, Jan 19, 2017 at 3:37 PM, Dylan Schiemann <[hidden email] >> <mailto:[hidden email]>> wrote: >> >> I see a few approaches: >> >> 1. Add to dojo/_base/lang. Probably the simplest, but limits what you >> can do. >> 2. Add a new dojo/mixin module. Ok, but would probably add confusion. >> 3. Add a new dojox/lang/mixin module. Probably the best place to add it, >> even though it might not get a lot of attention there. At least then you >> could add better versions of mixin and deepMixin and not risk confusion >> with what is in dojo/_base/lang >> >> So I'd probably lean towards #3 as the right approach given the >> constraints. >> >> Regards, >> -Dylan >> >> on 1/19/17, 11:18 (GMT-07:00) Michael Schall said the following: >> > I agree pulling in parts of Dojo 2 would be interesting, but probably >> > don't want to pull that much in for a single function... >> > >> > When I get time to look into pulling this back, would the correct end >> > goal be: >> > >> > Backward compatibility: only add a deepMixin (and renamed private >> called >> > functions as they collide) and leave current mixin (and called >> > functions) unchanged? >> > Smaller file: pull back both deepMixin and mixin? >> > >> > The smaller file approach would probably be best, but I'm not sure the >> > mixin functions between 1.x and 2.x are functionally equivalent. The >> > tests I can find for dojo 1.x lang.mixin >> > (https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js#L67 <https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js#L67> >> > <https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js >> <https://github.com/dojo/dojo/blob/master/tests/unit/_base/lang.js>>) >> > seem pretty sparse as compared to the ones for 2.x >> > (https://github.com/dojo/core/blob/master/tests/unit/lang.ts#L142 >> <https://github.com/dojo/core/blob/master/tests/unit/lang.ts#L142>). >> > >> > >> > On Thu, Jan 19, 2017 at 6:44 AM Dylan Schiemann >> <[hidden email] <mailto:[hidden email]> >> > <mailto:[hidden email] <mailto:[hidden email]>>> wrote: >> > >> > Hi Mike, >> > >> > We could. I'd be happy to review a PR if someone wants to >> backport it. >> > >> > Another option of course could be to use @dojo/core within your >> Dojo 1 >> > app. It would be larger than just adding a function to >> > dojo/_base/lang.js, but @dojo/core is pretty small. I'd be >> curious to >> > hear how that experience goes if you're interested in that >> experiment. >> > You could basically transpile @dojo/core and then use >> @dojo/core/lang as >> > an AMD module. >> > >> > Regards, >> > -Dylan >> > >> > on 1/18/17, 17:31 (GMT-07:00) Michael Schall said the following: >> > > Dylan - Thanks for the quick response! >> > > >> > > I tested this today a bit and I think this may work for our >> cases. >> > > >> > > Looks like dojo 2's core/lang exports a deepMixin function ( >> > > https://github.com/dojo/core/blob/master/src/lang.ts#L165 >> <https://github.com/dojo/core/blob/master/src/lang.ts#L165> ) that will >> > > support this in the future and has special handling for >> arrays similar >> > > to the jQuery implementation? >> > > >> > > Could core/lang's deepMixin be back ported to 1.x's lang module? >> > > >> > > Mike >> > > >> > > On Wed, Jan 18, 2017 at 6:32 AM, Dylan Schiemann >> > <[hidden email] <mailto:[hidden email]> >> <mailto:[hidden email] <mailto:[hidden email]>> >> > > <mailto:[hidden email] <mailto:[hidden email]> >> <mailto:[hidden email] <mailto:[hidden email]>>>> wrote: >> > > >> > > Hi Michael, >> > > >> > > It's in an odd location, but I believe what you are looking >> > for is the >> > > same as dojo/request/util.deepCopy >> > > >> > > See https://jsfiddle.net/dylan/oja45dv8/ >> <https://jsfiddle.net/dylan/oja45dv8/> >> > > <https://jsfiddle.net/dylan/oja45dv8/ >> <https://jsfiddle.net/dylan/oja45dv8/>> >> > > >> > > Regards, >> > > -Dylan >> > > >> > > on 1/17/17, 21:06 (GMT-07:00) Michael Schall said the >> following: >> > > > I opened an enhancement request >> > > > (https://bugs.dojotoolkit.org/ticket/18958 >> <https://bugs.dojotoolkit.org/ticket/18958> >> > > <https://bugs.dojotoolkit.org/ticket/18958 >> <https://bugs.dojotoolkit.org/ticket/18958>>) today but >> > probably should >> > > > have started here first... >> > > > >> > > > Basically I would like a deep mixin similar to jQuey's >> > extend(true, >> > > > obj1, obj2) overload... >> > > > >> > > > var object1 = { >> > > > apple: 0, >> > > > banana: { weight: 52, price: 100 }, >> > > > cherry: 97 >> > > > }; >> > > > var object2 = { >> > > > banana: { price: 200 }, >> > > > durian: 100 >> > > > }; >> > > > >> > > > // Merge object2 into object1, recursively >> > > > $.extend( true, object1, object2 ); >> > > > >> > > > //object1 = >> > > > >> > > >> > >> {"apple":0,"banana":{"weight":52,"price":200},"cherry":97,"durian":100} >> > > > >> > > > I have made a sample Plunker with jQuery.extend brute >> force >> > > converted to >> > > > lang.jmixin using dojo's isObject, IsArray and >> isFunction in >> > > app/lang.js >> > > > which is a copy of current dojo/_base/lang. It works for >> > the above >> > > > input, but as dojo doesn't have a function similar to >> > > isPlainObject that >> > > > I can see, it may not work exactly as theirs does. >> > > > >> > > > Plunker link: >> > http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview >> <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview> >> > > <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview >> <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview>> >> > > > >> > > -- >> > > Dojo Toolkit: http://dojotoolkit.org/ >> > > Tutorials: http://dojotoolkit.org/documentation/ >> <http://dojotoolkit.org/documentation/> >> > > <http://dojotoolkit.org/documentation/ >> <http://dojotoolkit.org/documentation/>> >> > > >> > > [hidden email] >> <mailto:[hidden email]> >> > <mailto:[hidden email] >> <mailto:[hidden email]>> >> > > <mailto:[hidden email] >> <mailto:[hidden email]> >> > <mailto:[hidden email] >> <mailto:[hidden email]>>> >> > > To unsubscribe, visit: >> > > >> http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest >> <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> >> > > >> <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest >> <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest>> >> > > >> > > >> > -- >> > Dojo Toolkit: http://dojotoolkit.org/ >> > Tutorials: http://dojotoolkit.org/documentation/ >> <http://dojotoolkit.org/documentation/> >> > >> > [hidden email] >> <mailto:[hidden email]> >> > <mailto:[hidden email] >> <mailto:[hidden email]>> >> > To unsubscribe, visit: >> > http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest >> <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> >> > >> -- >> Dojo Toolkit: http://dojotoolkit.org/ >> Tutorials: http://dojotoolkit.org/documentation/ >> <http://dojotoolkit.org/documentation/> >> >> [hidden email] >> <mailto:[hidden email]> >> To unsubscribe, visit: >> http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest >> <http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest> >> >> Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest
Co-Founder, Dojo Toolkit
CEO, SitePen, Inc. http://www.sitepen.com/ |
|
|
Dylan -
I got pulled onto another project so I haven't looked into pulling in the dojo2 modules. We did find an issue with the dojo1 request/util/deepCopy. It does not seem to support Date objects. I have updated the plunker (http://plnkr.co/edit/ Index: request/util.js ============================== --- request/util.js (revision 5861) +++ request/util.js (working copy) @@ -14,7 +14,11 @@ sval = source[name]; if(tval !== sval){ if(tval && typeof tval === 'object' && sval && typeof sval === 'object'){ - exports.deepCopy(tval, sval); + if(sval instanceof Date){ + target[name] = new Date(sval); + }else{ + exports.deepCopy(tval, sval); + } }else{ target[name] = sval; } Let me know what you think? Do I open a ticket for this? Mike -- Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest |
|
|
That sounds like a reasonable patch... If you open a ticket and a PR I
can land that pretty quickly (we're aiming for updated point releases in early March). Regards, -Dylan on 2/21/17, 07:06 (GMT-07:00) Michael Schall said the following: > Dylan - > > I got pulled onto another project so I haven't looked into pulling in > the dojo2 modules. We did find an issue with the dojo1 > request/util/deepCopy. It does not seem to support Date objects. > > I have updated the plunker > (http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview > <http://plnkr.co/edit/ebYbu9nQ5m3rAbyQbMP1?p=preview>) showing a date > property in the object structure. We have also found a fix and have > included the patch below and in the plunker example: > > Index: request/util.js > =================================================================== > --- request/util.js(revision 5861) > +++ request/util.js(working copy) > @@ -14,7 +14,11 @@ > sval = source[name]; > if(tval !== sval){ > if(tval && typeof tval === 'object' && sval && typeof sval === 'object'){ > -exports.deepCopy(tval, sval); > +if(sval instanceof Date){ > +target[name] = new Date(sval); > +}else{ > +exports.deepCopy(tval, sval); > +} > }else{ > target[name] = sval; > } > > Let me know what you think? Do I open a ticket for this? > > Mike > Dojo Toolkit: http://dojotoolkit.org/ Tutorials: http://dojotoolkit.org/documentation/ [hidden email] To unsubscribe, visit: http://mail.dojotoolkit.org/mailman/listinfo/dojo-interest
Co-Founder, Dojo Toolkit
CEO, SitePen, Inc. http://www.sitepen.com/ |
«
Return to Dojo Toolkit
|
1 view|%1 views
| Free forum by Nabble | Edit this page |